Aerosol Sci Eng (2017) 1:78–92 DOI 10.1007/s41810-017-0009-x

ORIGINAL PAPER

Comparison of Aerosol Optical Properties Between Two Nearby Urban Sites in Beijing, China Jie Yu1 • Huizheng Che1 • Quanliang Chen2 • Hongbin Chen3 • Philippe Goloub4 Ke Gui1,2 • Yu Zheng5 • Hong Wang1 • Yaqiang Wang1 • Linchang An6 • Tianze Sun1 • Xiaoye Zhang1 • Renjian Zhang7 • Mingkai Dai8

•

Received: 13 January 2017 / Revised: 2 April 2017 / Accepted: 17 May 2017 / Published online: 1 June 2017 Ó Institute of Earth Environment, Chinese Academy Sciences 2017

Abstract This study compares the aerosol optical properties measured by CE318 sunphotometers at the Institute of Atmospheric Physics and the Chinese Academy of Meteorological Sciences in Beijing between January 2013 and July 2015 to provide the framework to quantify the spatial and temporal variability of aerosol properties. Aerosol optical parameters included extinction (scattering plus absorption) aerosol optical depth (EAOD), extinction ˚ ngstro¨m exponent (EAE), columnar water vapor A (CWV), absorption aerosol optical depth (AAOD), ˚ ngstro¨m exponent (AAE), extinction aerosol absorption A optical depth of fine particles (EAODf), extinction aerosol optical depth of coarse particles (EAODc), real parts of the refractive index (REFR), imaginary parts of the refractive index (REFI), single scattering albedo (SSA), asymmetry factor (ASYM), size distribution, and sphericity fraction. Comparison of aerosol optical properties using the simultaneous observations between two sites showed that correlation coefficients were larger than

or equal to 0.98 for EAOD, EAE, and CWV, but smaller than or equal to 0.90 for AAOD, REFR, REFI, and SSA; the percentage differences for most of the parameters were less than 2%, but for EAODf were relatively large, ranging from 4.35 to 6.45%; the mean size distributions simultaneously showed bi-modal patterns, with two peak volumes at the radii of 0.15 and 2.94 lm; two kinds of tri-peak model were apparent during the study period; a case of EAODs at 440 nm differing by more than 0.2 between the two sites reflected the effect of local aerosol pollution. Comparison of aerosol characterization inferred by absorption properties using all the inversion data showed that classification using EAE, AAE, and sphericity fraction indicated the main aerosol type was ‘‘U/I&BB’’ (urban/industrial and biomass-burning), accounting for 59.87 and 57.43%, respectively; the volume size distribution retrievals binned by AAE exhibited coarse mode particles became dominant as AAE increased to 2.0; the SSA retrievals binned by AAE demonstrated

& Huizheng Che [email protected]

5

Key Laboratory for Aerosol-Cloud-Precipitation of China Meteorological Administration, Nanjing University of Information Science and Technology, Nanjing 210044, China

6

National Meteorological Center, CMA, Beijing 100081, China

7

Key Laboratory of Regional Climate-Environment Research for Temperate East Asia, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China

8

Key Laboratory for Semi-Arid Climate Change of the Ministry of Education, College of Atmospheric Sciences, Lanzhou University, Lanzhou 730000, China

1

2

State Key Laboratory of Severe Weather (LASW) and Institute of Atmospheric Composition, Chinese Academy of Meteorological Sciences CMA, Beijing 100081, China Plateau Atmospheric and Environment Key Laboratory of Sichuan Province, College of Atmospheric Sciences, Chengdu University of Information Technology, Chengdu 610225, China

3

Laboratory for Middle Atmosphere and Global Environment Observation, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029, China

4

Laboratoire d’Optique Amosphe´rique, Universite´ des Sciences et Technologies de Lille, 59655 Villeneuve d’Ascq, France

123

Aerosol Sci Eng (2017) 1:78–92

SSA transitioned from spectra representing dust to U/I&BB pollution; averaged SSA for all the retrievals and SSA data partitioned by the EAE and g675 nm suggested there were more absorbing aerosols at IAP. The results of the study will be beneficial in validating satellite observations and model simulation results, providing more accurate input parameters for model simulations. Keywords Comparison  Aerosol optical parameters  Absorption  Beijing

1 Introduction The impact of aerosol particles on climate is twofold (Aoki and Fujiyoshi 2003). On one hand, aerosols directly affect the earth–atmosphere radiative balance through scattering and absorbing shortwave and longwave radiations in the air (Charlson et al. 1992; IPCC 2013). On the other hand, indirectly, they serve as cloud condensation nuclei and iceforming nuclei to affect the size distribution of cloud droplets (Twomey et al. 1984). Many studies have emphasized that the nature of aerosol optical properties is one of the largest sources of uncertainty in current assessments of climate forcing (Ramanathan et al. 2001). Precise and consistent measurements of aerosol optical properties, such as EAOD (extinction aerosol optical depth), SSA (single scattering albedo), ASYM (asymmetry factor), and phase function, key parameters of the aerosol direct effect, are required in a timely manner to reduce such large uncertainties (Xie et al. 2011). Myhre (2009) also emphasized that research on the influence of aerosols needs more information on aerosol characteristics, such as size distribution, chemical composition and optical properties, especially light absorption properties. The power of aerosol optical properties in assessing and predicting global climate change needs further verification (Hansen et al. 2000). Detailed and accurate knowledge of aerosol optical properties remains insufficient (Aoki and Fujiyoshi 2003). EAOD and EAE can be used to study the size and growth of aerosol particles. The SSA defined as the ratio between the particle scattering coefficient and the total extinction coefficient is usually used to characterize the aerosol absorption properties and is a key variable in assessing the radiative forcing of aerosols (Bi et al. 2011). The SSA is mainly dependent on the chemical composition, size distribution and matter concentration of aerosol particles. The aerosol volume size distribution is one of the decisive factors in the transport and persistence of aerosols, as well as their optical properties. The ASYM represents an estimation of the asymmetry distribution of the scattering radiation. Currently, studies on aerosol optical properties focus mainly on EAOD and EAE (Xiao et al. 2011; Dong et al.

79

2013; Wang et al. 2010a; Che et al. 2011; Che et al. 2015a; Zhao et al. 2013), with relatively less research taking place on other aerosol optical parameters. Dubovik et al. (2000) successfully retrieved aerosol characteristics (size distribution, complex refractive index, and SSA), providing the inversion included the combined data of spectral optical depth together with sky radiances in the full solar almucantar-used radiances measured by ground-based Sun–sky scanning radiometers of the Aerosol Robotic Network (AERONET). Bi et al. (2011) discussed the aerosol optical properties according to a CIMEL sunphotometer of the Aerosol Robotic Network at the Semi-Arid Climate and Environment Observatory of Lanzhou University. Che et al. (2007) analyzed the seasonal variation of aerosol optical properties including EAOD, EAE, SSA, volume size distribution used skyradiometer observation data in Beijing. Eck et al. (2005) investigated the column-integrated optical properties of aerosols in the central eastern region of Asia and the tropical mid-Pacific based on Sun– sky radiometer measurements made at AERONET sites in these regions. Toledano et al. (2011) analyzed the variations in the EAOD in the range of 340–1550 nm, the ˚ ngstro¨m exponent, volume size distributions and single A scattering albedo during the SAMUM-2 experiment in the Cape Verde islands from January to February 2008. Because of the complexity of local emission sources in urban areas, model simulations and satellite inversions tend to be based on just one site, used as representative of the whole Beijing urban areas. This present study aims to use ground-based sunphotometer measurements from two neighboring urban AERONET sites in Beijing to investigate the influence of local pollution and the differences in aerosol optical properties in neighboring regions. Groundbased observation which is considered to be the most accurate aerosol research means is the basis of aerosol research; it can be obtained by various aerosol physical and radiative characteristics directly. The two nearby sites provided the framework to quantify the spatial and temporal variability of aerosol properties. Thus, more groundbased measurements of aerosol characteristics offered better representation to validate aerosol retrievals from satellites, improve satellite remote-sensing algorithms, and further improve the quality of satellite products. For the model simulation, the quantified spatial and temporal variability of aerosol properties will be useful for providing more accurate input parameters for the parametric scheme. The remainder of the paper is structured as follows: The study sites, instruments, and data are described first. Next, the averages of different aerosol optical parameters for simultaneous observations at the two urban sites in Beijing are compared and discussed. Third, cases in which three peaks in the size distribution and measurements of EAOD differed by more than 0.2 for simultaneous observations

123

80

between the two sites are then analyzed. Then, the aerosol characterizations inferred by absorption properties for all the inversion data are then analyzed. Finally, conclusions are drawn.

2 Sites, Instruments, Data and Method 2.1 Sites Figure 1 shows the two nearby (6 km apart) AERONET sites in Beijing: the Institute of Atmospheric Physics [IAP; (39.98°N, 116.38°E); 92.0 m ASL], Chinese Academy of Sciences and the Chinese Academy of Meteorological Sciences [CAMS; (39.93°N, 116.32°E); 106.0 m ASL], China Meteorological Administration. IAP is located between the north 3rd and 4th ring road in Beijing. The sampling site is located on a terrace on the roof of the IAP building (30 m high) in Beijing, which is next to the traffic road. While CAMS’s site is located in the interior of China Meteorological Administration, and away from the roadside. 2.2 Instruments Two CIMEL CE318 sunphotometers (one at each site) were installed in March 2001 at IAP and August 2012 at CAMS, and since then have been running continuously.

Fig. 1 Location of the IAP and CAMS sites in Beijing

123

Aerosol Sci Eng (2017) 1:78–92

The instruments differ in terms of the wavelengths at which they operate, but only the measurements of common channels (440, 675, 870, 1020 and 940 nm) were used in this study. The main components of the CE318 sunphotometer are the robot, optical head, collimator, and control box. The instrument takes direct spectral solar irradiance and sky radiance measurements for the solar almucantar scenario or principal plane scenario, within a 1.2° full fieldof-view every 15 min (Holben et al. 1998). Measurements of direct solar radiation were used to calculate the EAOD and the total amount of atmospheric columnar water vapor (CWV). Sky radiance measurements for the solar almucantar scenario or principal plane scenario can be used for the inversion of other aerosol optical parameters, such as SSA, size distribution, refractive index, ASYM, and phase function (Dubovik et al. 2000). The CIMEL sunphotometers of both IAP and CAMS are calibrated using the PHOTONS (http://www-loa.univlille1.fr/photons/) calibration facility in Lille, France, and at the Izan˜a Observatory, Spain, following the calibration protocol used by NASA, periodically. And the sunphotometer calibrations were conducted in August of each year at IAP and CAMS during the measurement period. The calibration procedure has been described by Che et al. (2009). The uncertainty in EAOD under a cloud-free condition is \±1% for wavelengths [440 nm and \±2% for shorter wavelengths (Eck et al. 1999). The uncertainty in

Aerosol Sci Eng (2017) 1:78–92

retrieval of CWV is typically \12% (Holben et al. 1998). The SSA with an expected accuracy to ±3% can be retrieved only for high aerosol loading (EAOD440 nm C0.40) for solar zenith angles [50° (Dubovik et al. 2002). For the intermediate particle size range (0.1 B r B 0.7 lm), the retrieval errors of particle volume size distribution do not exceed 10% in the maxima of the size distribution and may increase up to 35% for the points corresponding to the minimum values of dV(r)/dln r within this size range; for the particles less than 0.1 lm and larger than 0.7 lm, the accuracy decreases significantly because of the low sensitivity of the aerosol scattering at 440, 675, 870 and 1020 nm to particles in these size ranges (Dubovik et al. 2002; Dubovik and King 2000; Dubovik et al. 2000). The errors are estimated to be ±4% in the real parts and 30–50% in the imaginary parts of the refractive index (Dubovik et al. 2000). An accuracy assessment of the AERONET retrievals can be found in the work of Dubovik et al. (2000, 2006). We employ AERONET Level 1.5 data in this study which are quality assured and cloud screened by the method of Smirnov et al. (2000). 2.3 Data and Method AERONET data from January 2013 to July 2015 are used in this paper (http://aeronet.gsfc.nasa.gov/). For details regarding the version 2 algorithm, readers are referred to Dubovik et al. (2006) and Eck et al. (2008). The aerosol optical parameters in the AERONET database include EAOD, EAE, CWV, absorption AOD (AAOD), extinction AOD of fine particles (EAODf), extinction AOD of coarse particles (EAODc), real parts of the refractive index (REFR), imaginary parts of the refractive index (REFI), SSA, ASYM, volume size distribution, and sphericity fraction (fraction of spherical to spheroidal plus spherical particles). These parameters were used to compare the aerosol optical properties between IAP and CAMS. The aerosol inversions between the two sites in Beijing were compared using the data measured less than 1 min apart (can be seen as simultaneous observations). Comparisons of EAOD, EAE440–870 nm and CWV between IAP and CAMS were based on 5970 pairs of measurements taken within 1 min from each other for the 477 days of the study period. Only results with EAOD440 nm [0.40 were used to retrieve the SSA, because of the greater level of uncertainty inherent in lower EAOD values (Che et al. 2015b). Therefore, comparisons of other aerosol optical parameters between IAP and CAMS were based on 126 pairs of measurements taken within 1 min from each other on 75 days. Comparisons of aerosol characterization inferred by absorption properties for all the retrievals were based on 942 measurements at IAP and 707 measurements at CAMS.

81 Fig. 2 Scattergrams of aerosol optical parameters between IAP and c CAMS: a EAOD; b EAE; c CWV; d AAOD; e EAODf; f EAODc; g REFR; h REFI; i SSA; j ASYM

3 Results 3.1 Comparison of Aerosol Optical Properties for Simultaneous Observations Figure 2 shows the results that enable us to compare the aerosol optical parameters between IAP and CAMS. The light-blue dashed line in each panel of Fig. 2 is the 1:1 ideal line (i.e., y = x). Additionally, mean values and differences in aerosol optical parameters between IAP and CAMS are shown in Table 1. Figure 2a shows the EAOD values and linear fitting curve at each wavelength between the two sites. High correlation was found, with a correlation coefficient of 0.99 at each wavelength. By and large, EAOD values at IAP were slightly higher than those at CAMS, the difference that could be associated with the fact that IAP has a roadside location. The mean EAODs decreased with increasing wavelength. The mean EAOD440 nm at IAP and CAMS was 0.53 ± 0.56 (the number behind the symbol ‘‘±’’ means the standard deviation) and 0.52 ± 0.56, respectively. The percentage differences in EAOD between the two sites at 440, 675, 870 and 1020 nm, calculated using the formula EAOD  EAODkCAMS kIAP  100%; ð1Þ EAODkCAMS where 1.92, 0, 4.35 and 0%, respectively. This confirms the high consistency in EAOD for the two sets of CIMEL sunphotometer measurements at IAP and CAMS. The uncertainty in EAODs is less than ±1% for wavelengths [440 nm (Eck et al. 1999). Therefore, the large percentage differences of EAOD at 440 and 870 nm may be caused by the different aerosol properties at two sites. There was notable linear correlation between the EAEs of the two sites computed from the instantaneous measurements (Fig. 2b). The correlation coefficient between the EAEs at 440–870 nm was 0.98. The linear fitting curve of the EAEs was slightly higher than the y = x line. The mean EAE at IAP and CAMS was 1.12 ± 0.34 and 1.13 ± 0.35, respectively. The total average EAE based on 5970 pairs of data between the two sites differed by 0.88%. On the whole, the amount of water vapor at CAMS was higher than that at IAP (Fig. 2c), which could be due to the presence of the air conditioner outdoor unit beside the CAMS’s site and the presence of the lake around the CAMS’s site. Nevertheless, a highly significant linear relationship was found between the two sites, with a

123

82

Aerosol Sci Eng (2017) 1:78–92 5.0

EAOD440nm EAOD675nm EAOD870nm EAOD1020nm

4.5 4.0 3.5

2.0

a

b

EAE440-870nm linear fit of EAE R=0.98

1.5

CAMS

CAMS

3.0 2.5 2.0 1.5

1.0

0.5

1.0

linear fit of 440 R=0.99 linear fit of 675 R=0.99 linear fit of 870 R=0.99 linear fit of 1020 R=0.99

0.5 0.0

0.0

-0.5 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

0.0

0.5

IAP

6

c

CWV(cm) linear fit of CWV R=0.99

5

AAOD440nm AAOD675nm AAOD870nm AAOD1020nm

0.20 0.15

CAMS

4

CAMS

1.0

3 2

0.10 0.05 linear fit of 440 linear fit of 675 linear fit of 870 linear fit of 1020

0.00

0

1

2

3

4

5

0.00

6

0.05

0.10

3.5

EAODf440nm EAODf675nm EAODf870nm EAODf1020nm

3.0 2.5

0.15

R=0.90 R=0.82 R=0.75 R=0.77

0.20

IAP

IAP 0.8

e

0.7 0.6

EAODc440nm EAODc675nm EAODc870nm EAODc1020nm

f

0.5

2.0

CAMS

CAMS

2.0

d

1 0

1.5

IAP

1.5 1.0 linear fit of 440 R=0.98 linear fit of 675 R=0.98 linear fit of 870 R=0.97 linear fit of 1020 R=0.97

0.5 0.0 0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.4 0.3 0.2

linear fit of 440 R=0.97 linear fit of 675 R=0.97 linear fit of 870 R=0.97 linear fit of 1020 R=0.97

0.1 0.0 0.0

3.5

0.1

0.2

0.3

IAP 1.60 1.55

REFR 440nm REFR 675nm REFR 870nm REFR 1020nm

0.4

0.5

0.6

0.7

0.8

IAP 0.030

REFI 440nm REFI 675nm REFI 870nm REFI 1020nm

g 0.025 0.020

h

CAMS

CAMS

1.50 1.45

0.015 0.010

1.40 linear fit of 440 R=0.63 linear fit of 675 R=0.78 linear fit of 870 R=0.84 linear fit of 1020 R=0.85

1.35 1.30 1.30

1.35

1.40

1.45

1.50

1.55

linear fit of 440 R=0.89 linear fit of 675 R=0.84 linear fit of 870 R=0.82 linear fit of 1020 R=0.85

0.005 0.000

1.60

0.000

0.005

0.010

IAP 1.00 0.98 0.96

0.80

ASYM440nm ASYM675nm ASYM870nm ASYM1020nm

i 0.76 0.72

0.92

CAMS

CAMS

0.94

SSA440nm SSA675nm SSA870nm SSA1020nm

0.015

0.020

0.025

0.030

IAP

0.90 0.88

j

0.68 0.64

0.86 linear fit of 440 R=0.89 linear fit of 675 R=0.84 linear fit of 870 R=0.81 linear fit of 1020 R=0.83

0.84 0.82

0.80 0.80 0.82 0.84 0.86 0.88 0.90 0.92 0.94 0.96 0.98 1.00 IAP

123

linear fit of 440 R=0.84 linear fit of 675 R=0.90 linear fit of 870 R=0.91 linear fit of 1020 R=0.92

0.60 0.56 0.56

0.60

0.64

0.68 IAP

0.72

0.76

0.80

Aerosol Sci Eng (2017) 1:78–92

83

Table 1 Mean values and standard deviations of aerosol optical parameters at IAP and CAMS, and the differences in aerosol optical parameters between the two sites Parameter

IAP

CAMS

Difference

Parameter

IAP

CAMS

Difference

EAOD440

0.53 ± 0.56

0.52 ± 0.56

1.92%

EAODc1020

0.18 ± 0.13

0.18 ± 0.13

0

EAOD675

0.32 ± 0.35

0.32 ± 0.34

0

REFR440

1.47 ± 0.06

1.47 ± 0.06

0

EAOD870

0.24 ± 0.25

0.23 ± 0.25

4.35%

REFR675

1.49 ± 0.05

1.49 ± 0.05

0

EAOD1020

0.20 ± 0.20

0.20 ± 0.20

0

REFR870

1.49 ± 0.05

1.49 ± 0.05

0

EAE

1.12 ± 0.34

1.13 ± 0.35

0.88%

REFR1020

1.49 ± 0.05

1.50 ± 0.05

0.67%

CWV (cm)

1.09 ± 0.88

1.11 ± 0.95

1.80%

REFI440

0.01 ± 0.00

0.01 ± 0.01

0

AAOD440 AAOD675

0.09 ± 0.04 0.04 ± 0.02

0.09 ± 0.04 0.04 ± 0.02

0 0

REFI675 REFI870

0.01 ± 0.00 0.01 ± 0.00

0.01 ± 0.00 0.01 ± 0.00

0 0

AAOD870

0.03 ± 0.01

0.03 ± 0.01

0

REFI1020

0.01 ± 0.00

0.01 ± 0.00

0

AAOD1020

0.03 ± 0.01

0.03 ± 0.01

0

SSA440

0.92 ± 0.04

0.92 ± 0.04

0

EAODf440

1.00 ± 0.53

0.95 ± 0.47

5.26%

SSA675

0.94 ± 0.03

0.93 ± 0.03

1.08%

EAODf675

0.53 ± 0.33

0.50 ± 0.30

6.00%

SSA870

0.93 ± 0.03

0.93 ± 0.03

0

EAODf870

0.33 ± 0.22

0.31 ± 0.20

6.45%

SSA10200

0.93 ± 0.04

0.93 ± 0.04

0

EAODf1020

0.24 ± 0.17

0.23 ± 0.15

4.35%

ASYM440

0.71 ± 0.03

0.71 ± 0.03

0

EAODc440

0.15 ± 0.11

0.15 ± 0.11

0

ASYM675

0.66 ± 0.64

0.65 ± 0.03

1.54%

EAODc675

0.16 ± 0.12

0.16 ± 0.12

0

ASYM870

0.64 ± 0.03

0.63 ± 0.04

1.59%

EAODc870

0.17 ± 0.13

0.17 ± 0.12

0

ASYM1020

0.63 ± 0.03

0.63 ± 0.04

0

440–870

correlation coefficient of 0.99. The total average CWV based on 5970 pairs of data between the two sites differed by 1.80%. The correlation coefficients between the AAODs ranged from 0.75 to 0.90 (Fig. 2d). Relatively poor correlation may be due to the different kinds of aerosols caused by local emission source, especially absorption aerosol particles, such as the black carbon and brown carbon (Myhre 2009; Che et al. 2014). Song et al. (2013) pointed out the average black carbon concentrations at the roadside site in Beijing were 12.3 and 17.9 lg m-3 in summer and winter, respectively. The mean AAODs between two sites are the same at all four wavelengths, ranging from 0.03 ± 0.01 to 0.09 ± 0.04. When EAODf values were larger than 0.5, the EAODfs at IAP were larger than those at CAMS (Fig. 2e). The correlation coefficients for the EAODf at 440, 675, 870 and 1020 nm were 0.98, 0.98, 0.97, and 0.97, with slopes of 0.91, 0.89, 0.87 and 0.87, respectively. The mean EAODf decreased with increasing wavelength, ranging from 0.24 ± 0.17 (1020 nm) to 1.00 ± 0.53 (440 nm) at IAP and from 0.23 ± 0.15 (1020 nm) to 0.95 ± 0.49 (440 nm) at CAMS. The differences in EAODf ranged from 4.35% (1020 nm) to 6.45% (870 nm), which were the largest of the differences among all the measured parameters. The difference in EAODf was likely due to differences in the chemical components of aerosol (Zhang et al. 2013a, b), suggesting that the local aerosol contribution was distinct (Wang et al. 2010b).

When EAODc values were larger than 0.2, the results (Fig. 2f) were similar to those of EAODf. Furthermore, a highly significant linear relationship (0.97 at four wavelengths) between the two sites was found, indicating that the deviation between the two groups of data was small. The mean EAODc increased with increasing wavelength, but was smaller than its EAODf equivalent. This indicated that the fine mode aerosol particles were the dominant EAOD contributor. The mean EAODc values between two sites were the same at four wavelengths, ranged from 0.15 ± 0.11 to 0.18 ± 0.13. The linear fitting curve of REFR was higher than the y = x line at 1020 nm, with a strong correlation of R = 0.85, but the situation was more complicated at other three wavelengths, with the correlation coefficients ranging from 0.63 (440 nm) to 0.84 (870 nm) (Fig. 2g). The deviation between the two groups of data was larger at 440 nm. The aerosol properties at shorter wavelengths have much greater sensitivity due to high scattering properties (Eck et al. 1999). Therefore, the larger deviation at 440 nm may be due to the greater sensitivity at shorter wavelengths. The mean REFR values were the same at each wavelength between the two sites. Figure 2h shows that the scatter of REFI was relatively concentrated compared with the scatter of REFR. When REFIs were larger than 0.01, the linear fitting curves were lower than the y = x line, at all four wavelengths. The correlation coefficients between the REFI at 440, 675, 870 and 1020 nm were 0.89, 0.84, 0.82 and 0.85, with slopes of

123

84

Aerosol Sci Eng (2017) 1:78–92

0.86, 0.82, 0.82 and 0.85, respectively. The mean REFI values were 0.01 for four wavelengths at both IAP and CAMS. Similar to REFI, the SSA patterns were scattered with correlation coefficients ranging from 0.81 to 0.89 in the linear fit scatter plots (Fig. 2i). SSA is related to refractive index. The scatter patterns of REFR, REFI and SSA shows that the chemical components of the aerosol particles were different, suggesting that the impact of local emissions cannot be ignored. The mean SSA values were same at 440, 870 and 1020 nm, while a subtle difference of 1.08% was found at 675 nm. Figure 2j shows that correlation coefficients of ASYM are larger than 0.84. However, the linear fit pattern of ASYM at 440 nm is scattered, with a correlation coefficient and slope of 0.84 and 0.81, respectively. For symmetrical scattering, Rayleigh scattering, the ASYM is considered to be 0, and for a purely forward scattering aerosol the value is taken as 1. For a cloudless atmosphere, the ASYM ranges from 0.1 in very clean conditions to 0.75 in polluted conditions (Zege et al. 1991). In the present study, the mean ASYMs ranged from 0.63 ± 0.03 (1020 nm) to 0.71 ± 0.03 (440 nm) and from 0.63 ± 0.04 (1020 nm) to 0.71 ± 0.03 (440 nm) at IAP and CAMS, respectively. The differences between the two sites were less than 2%, at four wavelengths. Figure 3 shows the mean aerosol volume size distributions with standard deviation at IAP and CAMS. The mean values differed little between the two sites. One can see that the size distributions simultaneously showed a bimodal logarithm normal structure: fine mode with radius \0.6 lm and a coarse mode with radius [0.6 lm (Dubovik et al. 2002). As shown in Fig. 3, two peak volumes are at radii of 0.15 and 2.94 lm and with volume size spectra (dV/dln r) of 0.09 ± 0.05 and 0.12 ± 0.07 lm3/lm2 at IAP and 0.09 ± 0.04 and 0.11 ± 0.07 lm3/lm2 at CAMS. Thus, it 0.20 0.18 0.16

IAP CAMS

2

0.12

3

dV/dlnr(μm /μm )

0.14 0.10 0.08 0.06 0.04 0.02 0.00 -0.02

0.1

1

10

Radius(μm)

Fig. 3 Mean aerosols volume size distributions with standard deviation at IAP and CAMS

123

can be inferred that the extinction at Beijing could be associated with fine mode aerosols and coarse mode aerosols. The volume concentration of coarse particles in Beijing was higher than those of Mexico, Paris, and Hangzhou, China, but the volume concentration of fine particles were similar (Dubovik et al. 2002; Qi et al. 2016). This means that there were more coarse aerosol loading in Beijing among these urban sites (Zhang et al. 2012). The mean effective radius of the fine mode was 0.17 ± 0.03 and 0.16 ± 0.03 lm at IAP and CAMS, respectively. The mean effective radius of the coarse mode was 2.25 ± 0.34 and 2.21 ± 0.30 lm, respectively. The differences in effective radius of the fine mode and coarse mode between the two sites were 6.25 and 1.81%, respectively. The larger difference in effective radius of the fine mode may be due to the different physical and chemical properties of fine mode particles at two closed sites. 3.2 Aerosol Volume Size Distributions with a TriModal Pattern Aerosol volume size distributions with a tri-modal pattern were picked out from 126 data pairs at the two sites, as shown in Fig. 4. There were two peaks at the radii of about 0.10–0.20 and 0.30–0.50 lm for the fine mode and one peak at a radius of about 1.0–4.0 lm for the coarse mode in January and April 2013 (Fig. 4a–c, e). According to Che et al. (2014), 10–16 January, 2013 was the most intense haze period and the relative humidity was more than 60%. This phenomenon that occurred in winter and spring probably reflected the hygroscopic characteristics of fine particles (Che et al. 2014). According to the results reported by Zhang et al. (2013a), the mean measured nonrefractory submicron particle mass concentration at IAP is composed of organics (49.8%), sulphate (21.4%), nitrate (14.6%), ammonium (10.4%), and chloride (3.8%) in January 2013. These hygroscopic compositions could be one of the major reasons for the tri-modal size distribution. Another possible reason of the bi-mode submicron size distribution could be associated with the haze-fog and cloud processing (Li et al. 2011; Eck et al. 2012). Because the distance between IAP and CAMS is less than 6 km, CAMS had a similar tri-modal size distribution. And Che et al. (2014) also found the similar tri-modal pattern at CAMS during the same period. However, there were one peak at a radius of about 0.10–0.15 lm for the fine mode and two peaks at radii of about 1.30–1.80 and 3.80–3.90 lm for the coarse mode in April and May 2013 (Fig. 4d, f). On 22 May 2013, visibility was good, wind speed was low, and there was no haze. However, the CWV was 1.84 cm at IAP and 2.09 cm at CAMS at the moment featured in Fig. 4d and f. Therefore, the phenomenon of one peak for the fine mode and two peaks for the coarse mode was

Aerosol Sci Eng (2017) 1:78–92

85

11/1/2013 14:24:19 IAP CAMS 11/1/2013 14:25:02

a 2

3

0.22 0.20 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00

dV/dlnr( μ m /μm )

3

2

dV/dlnr( μm /μm )

Fig. 4 Aerosol volume size distributions with a tri-modal pattern at IAP and CAMS

0.1

1

10

0.22 0.20 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00

14/1/2013 10:25:27 IAP CAMS 14/1/2013 10:25:05

0.1

14/1/2013 11:25:26 IAP CAMS 14/1/2013 11:25:52

c 2

3

0.22 0.20 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00

0.1

1

10

0.22 0.20 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00

0.1

e 2

3

0.1

1 Radius(μm)

associated with the coagulation and hygroscopic growth of aerosol particles. 3.3 A case of Instantaneous EAOD440 nm Differing by More than 0.2 Between the Two Sites A case (on 23 June 2013) in which the EAODs at 440 nm differed by more than 0.2 for all the instantaneous data between the two sites (|EAODIAP - EAODCAMS| [0.2), including the corresponding CWV and EAE values, is analyzed. Figure 5a shows the daily averaged EAOD440 nm, CWV and EAE at two sites for the day. And the instantaneous data of the three parameters are also displayed in Fig. 5b, c and d, respectively. The daily averaged values are 0.84 and 0.67 for EAOD440 nm, 2.68 and 3.00 for CWV, and 1.35 and 1.60 for EAE at IAP and CAMS, respectively.

d

1

10

Radius(μm)

dV/dlnr( μm /μm )

3

2

dV/dlnr( μ m /μ m )

24/4/2013 8:09:46 IAP CAMS 24/4/2013 8:10:07

10

3/4/2013 9:16:47 IAP CAMS 3/4/2013 9:17:04

Radius(μm) 0.22 0.20 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00

1 Radius(μm)

dV/dlnr( μ m /μm )

3

2

dV/dlnr( μm /μm )

Radius(μm)

b

10

0.22 0.20 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00

22/5/2013 8:15:46 IAP CAMS 22/5/2013 8:16:05

0.1

f

1

10

Radius(μm)

This indicates that the daily averaged EAOD440 nm of IAP was larger than that of CAMS and the aerosol particles at IAP were larger than those at CAMS. This could have been associated with the fact that IAP is located at the side of a road, and the different chemical composition of aerosols (Zhang et al. 2013a). The surface-observed data and weather conditions on 23 June 2013 indicate that visibility was good, wind speed was low and there was no haze. 3.4 Aerosol Characterization Inferred by Absorption Properties for All Inversions EAE is a good indicator of aerosol particle size, while AAE is an intrinsic property closely related to aerosol chemical components (Giles et al. 2012). The fine mode fraction of EAOD [g = EAODf/(EAODf ? EAODc)] is also an

123

Aerosol Sci Eng (2017) 1:78–92 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

1.4

IAP CAMS

June 23, 2013

a

IAP CAMS

1.2

b

1.0 0.8

EAOD

value

86

0.6 0.4 0.2 0.0

EAOD 440nm

CWV(cm)

7:10

AE

9:17

9:32 10:17 10:47 14:17 14:47 16:24 17:11 17:43 17:51

Time

4.0

2.0 IAP CAMS

3.5

c

IAP CAMS

1.8

d

1.6 3.0 1.4 1.2

EAE

CWV(cm)

2.5 2.0 1.5

1.0 0.8 0.6

1.0 0.4 0.5 0.0

0.2 7:10

9:17

9:32 10:17 10:47 14:17 14:47 16:24 17:11 17:43 17:51

Time

Fig. 5 Daily mean EAOD440 CAMS

nm,

7:10

9:17

9:32 10:17 10:47 14:17 14:47 16:24 17:11 17:43 17:51

Time

CWV and EAE and all simultaneous instantaneous data for the three parameters on 23 June 2013 at IAP and

aerosol size proxy. According to Giles et al. (2011, 2012), the relationship of aerosol absorption and size information [e.g., EAE or g675 nm (g at 675 nm)] could be used to categorize the dominant aerosol type or optical mixture. Therefore, both the relationships of instantaneous AAE vs. EEA (hereafter defined as ‘‘AAE/EAE’’) and AAE vs. g675 nm (hereafter defined as ‘‘AAE/g675 nm’’) were analyzed with AERONET Version 2, Level 1.5 EAOD and almucantar retrievals for IAP and CAMS in Fig. 6. The green ellipses denote commonly used aerosol classifications: dust, mixed, and U/I&BB (urban/industrial and biomass-burning). The values of AAE, EAE, and g675 nm for different aerosol types were selected based on the literature (Giles et al. 2011, 2012) and characteristics of the stations. For IAP and CAMS, the AAE/EAE and AAE/ g675 nm relationships show a nonlinear dependence over the aerosol size ranges, whereas the sphericity fraction has a strong transition from non-spherical to spherical particles around EAE of *1.3 or g675 nm of *0.66. The ‘‘Mostly Dust’’ category is characterized by EAE B 0.5 (g675 nm B0.33), sphericity fraction \0.2 and AAE [2.0. The

123

0.0

EAE [ 0.8 (g675 nm [0.66), and sphericity fraction C0.2 associated with 1.0 \ AAE B 2.0 represents the ‘‘U/ I&BB’’ aerosol type. The ‘‘Mixed’’ category is classified as 0.5 \ EAE \ 0.8 (0.33 \ g675 nm \ 0.66) and 1.0 \ AAE B 2.0. For the AERONET data, AAE \1.0 may be related to black carbon coating by coarse particles such as dust and organic carbon (Dubovik et al. 2000; Eck et al. 2010). At IAP (Fig. 6a) and CAMS (Fig. 6b), the main aerosol type is ‘‘U/I&BB’’, accounting for 59.87 and 57.43%, respectively; and the ‘‘Mostly Dust’’ category accounting for 2.12 and 1.56%, respectively. Such a large number of fine mode aerosol loading is associated with cooking emission, traffic emissions, sporadic coal combustion emissions, nearby industrial emissions, and biomass-burning (Sun et al. 2016; Zhang et al. 2013b). The volume size distribution retrievals were binned based on AAE in Fig. 7. The size distributions presented bi-modal patterns for both IAP and CAMS: fine mode (a peak between 0.1 and 0.2 lm) and coarse mode (a peak between 2 and 4 lm). As AAE increased to 2.0, coarse mode particles became dominant.

3.0 Mixed 2.5

U/I & BB

Sphericity Fraction

2.0 1.5 1.0 0.5 0.0 -0.5 0.0 0.5 1.0 1.5 Extinction Ångström Exponent (440-870 nm)

3.5

IAP

Mostly Dust

3.0 2.5

Mixed

c

2.0

AAE Sphericity

U/I & BB

2.0 1.5 1.0 0.5 0.0 0.0 0.2 0.4 0.6 0.8 1.0 Fine Mode Fraction of Aerosol Optical Depth (675 nm)

3.5

CAMS

b

AAE Sphericity

Mostly Dust

3.0 Mixed 2.5

U/I & BB 2.0

Sphericity Fraction

AAE Sphericity

Mostly Dust

1.5 1.0 0.5 0.0 -0.5 0.0 0.5 1.0 1.5 2.0 Extinction Ångström Exponent (440-870 nm)

3.5

CAMS

Mostly Dust

3.0 2.5

Mixed

d

AAE Sphericity

U/I & BB

2.0

Sphericity Fraction

a

Absorption Ångström Exponent (440-870 nm)

IAP

87

Absorption Ångström Exponent (440-870 nm)

3.5

Sphericity Fraction

Absorption Ångström Exponent (440-870 nm)

Absorption Ångström Exponent (440-870 nm)

Aerosol Sci Eng (2017) 1:78–92

1.5 1.0 0.5 0.0 0.0 0.2 0.4 0.6 0.8 1.0 Fine Mode Fraction of Aerosol Optical Depth (675 nm)

˚ ngstro¨m exponent (AAE) and sphericity Fig. 6 The absorption A ˚ ngstro¨m exponent (EAE) and fraction as a function of extinction A fine mode fraction of EAOD at 675 nm (g675 nm; from the almucantar inversions) using AERONET version 2, Level 1.5 data of IAP and CAMS. AAE is plotted from 0.0 to 3.5 (red) and sphericity fraction is

plotted from 0.0 to 1.0 (blue). The green ellipses represent probable aerosol mixture categories. AAE of 1.0 indicates k-1 dependence, and a sphericity fraction of 1.0 indicates a 100% spherical particle (color figure online)

The SSA retrievals binned by AAE were depicted in Fig. 8. And the two urban closed sites showed the similar variation trends. SSA transitioned from spectra representing dust (i.e., strong absorption in the blue wavelength region and relatively weak absorption in the near-infrared) to U/I&BB pollution (i.e., stronger absorption in longer wavelengths); the interpretation of these SSA spectra is consistent with results reported by Giles et al. (2011), Prasad and Singh (2007) and Derimian et al. (2008). However, as AAEs decrease to 1.25, the mean SSA values at IAP were less than those at CAMS for all wavelengths. AAE *1.0 is often associated with black carbon, either with or without coatings (Gyawali et al. 2009). It suggests that there were more black carbon aerosols at IAP which is a roadside site (Song et al. 2013; Yan et al. 2014). Figure 9 demonstrated mean SSA for all the retrievals and SSA data

were further partitioned based on the EAE intervals of 0.0–0.8 and 0.8–2.0 and g675 nm intervals of 0.0–0.33, 0.33–0.66, and 0.66–1.0 at two sites. Figure 10 showed the mean AAOD, mean EAODf and mean EAODc for all retrievals at two sites. Strong absorption was noted at 440 nm relative to longer wavelengths for all SSA retrievals (Fig. 9a), SSA with EAE [0.8 (Fig. 9c), and SSA with g [ 0.66 (Fig. 9f), indicating aerosols were dominated by fine mode particles both at IAP and CAMS. However, the SSA values at IAP were lower than those at CAMS for the cases of the above, suggesting there are more absorbing aerosol particles at IAP (Song et al. 2013; Yan et al. 2014). The lower SSA values at IAP were consistent with the higher AAODs and higher EAODfs (Fig. 10). Strong absorption at 440 nm and strong scattering at longer wavelengths for SSA with AAE B0.8

123

Aerosol Sci Eng (2017) 1:78–92

IAP

2 3

0.30 Binned byAAE 0.28 0.0-1.0 n=64 0.26 1.0-1.25 n=158 0.24 1.25-1.5 n=222 0.22 1.5-2.0 n=379 0.20 2.0-2.5 n=93 >2.5 n=26 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 0.01 0.1

dV(r)/dln(r) (μm /μm )

3

2

dV(r)/dln(r) (μ m /μm )

88

Radius (μm)

1

10

0.30 Binned byAAE 0.28 0.0-1.0 n=68 0.26 1.0-1.25 n=129 0.24 1.25-1.5 n=137 0.22 1.5-2.0 n=307 2.0-2.5 n=51 0.20 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 0.01 0.1

CAMS

1

10

Radius (μ m)

Fig. 7 Aerosol volume size distribution by AAE bins for IAP and CAMS. Averages in which n \ 25 were removed from the plots (n represents the sample size) 1.00

1.00

CAMS

Single Scattering Albedo

Single Scattering Albedo

IAP

0.95

0.90

Binned byAAE 0.0-1.0 1.0-1.25 1.25-1.5 1.5-2.0 2.0-2.5 >2.5

n=64 n=158 n=222 n=379 n=93 n=26

0.85 400

500

600

700

800

900

1000

1100

0.95

0.90

Binned byAAE 0.0-1.0 1.0-1.25 1.25-1.5 1.5-2.0 2.0-2.5

0.85 400

Wavelength (nm)

500

600

n=68 n=129 n=137 n=307 n=51

700

800

900

1000

1100

Wavelength (nm)

Fig. 8 SSA averaged by AAE bins for IAP and CAMS. Averages in which n \ 25 were removed from the plots

(Fig. 9b) and SSA with g \ 0.33 (Fig. 9d) represented large dust particles (Giles et al. 2011). And the higher SSA values at IAP except at 440 nm were consistent with the higher EAODcs (Fig. 10). This may be associated with the dust absorbing particles and mixed particles.

4 Conclusions The aerosol optical properties were compared and discussed at two neighboring urban sites of Beijing between January 2013 and July 2015, including optical parameters for simultaneous observations and absorption properties for all the inversions. The study yielded the following conclusions:

123

(1)

High correlations were found, with correlation coefficients larger than or equal to 0.98, for the ˚ ngextinction aerosol optical depth, extinction A stro¨m exponent, columnar water vapor; correlation coefficients were smaller than or equal to 0.90 for absorption aerosol optical depth, real parts of the refractive index, imaginary parts of the refractive index, and single scattering albedo. The percentage differences were almost less than 2.0%, but greater than 4.0% for extinction aerosol optical depth at 675 nm and extinction aerosol optical depth of fine particles at four wavelengths. The mean volume size distributions simultaneously showed bi-modal patterns, with two peaks at radii of 0.15 and 2.94 lm and with volume size spectra (dV/dln r) of

Aerosol Sci Eng (2017) 1:78–92

89

1.00

0.96

0.96

0.96

0.94

0.94

0.94

0.98

SSA

EAE ≤0.8

0.92

0.92

0.92

0.90

0.90

0.90

0.88 440nm

1.00 0.98

675nm

870nm

440nm 1.00

η≤0.33

675nm

870nm

440nm

1020nm

0.33<η≤0.66

0.94

0.94

0.94

SSA

0.96

SSA

0.96

0.92

0.92

0.92

0.90

0.90

0.90

0.88 440nm

675nm

870nm

1020nm

a

675nm

870nm

1.8 1.6

440nm

675nm

870nm

1020nm

b

IAP CAMS

1.2 EAOD-F

0.10 AAOD

1020nm

1.4

0.12

0.08 0.06

1.0 0.8 0.6

0.04

0.4

0.02

0.2

0.00

0.0 440nm

0.40

f

η>0.66

conditions (i.e., EAE is [0.8); d mainly coarse mode particles (g675 nm B 0.33); e mixed size particles (0.33 \ g675 nm B 0.66); and f mainly fine mode particles (g675 nm [ 0.66)

IAP CAMS

0.14

1020nm

0.88 440nm

Fig. 9 Mean SSA and SSA data were partitioned based on EAE and g675 nm using IAP and CAMS AERONET. a Mean SSA for all the SSA retrievals at two sites. b The case for large particle-dominated conditions (i.e., EAE is B0.8); c the case for small particle-dominated 0.16

870nm

IAP CAMS

0.98

0.96

0.88

675nm

1.00

e

IAP CAMS

d 0.98

EAE>0.8

0.88

1020nm

IAP CAMS

c

IAP CAMS

b

0.98

0.88

SSA

1.00

IAP CAMS

0.98

SSA

SSA

1.00

a

IAP CAMS

675nm

870nm

c

1020nm

440nm

675nm

870nm

1020nm

IAP CAMS

0.35 0.30

EAOD-C

0.25 0.20 0.15 0.10 0.05 0.00 440nm

675nm

870nm

1020nm

Fig. 10 a Mean AAOD for all retrievals, b mean EAODf for all retrievals, c mean EAODc for retrievals at two sites

123

90

(2)

(3)

Aerosol Sci Eng (2017) 1:78–92

0.09 ± 0.05 and 0.12 ± 0.07 at IAP and 0.09 ± 0.04 and 0.11 ± 0.07 at CAMS. Cases of two kinds of tri-peak model were analyzed during the study period: one with two peaks at radii of about 0.10–0.20 and 0.30–0.50 lm for the fine mode and one peak at a radius of about 1.0–4.0 lm for the coarse mode in January and April 2013, possibly reflecting the hygroscopic characteristics of fine particles; and another with one peak at a radius of about 0.10–0.15 lm for the fine mode and two peaks at radii of about 1.30–1.80 and 3.80–3.90 lm for the coarse mode in April and May 2013, probably due to coagulation and hygroscopic growth of aerosol particles. A case when extinction aerosol optical depths at 440 nm differed by more than 0.20 for simultaneously instantaneous data between the two sites reflected the effect of local aerosol pollution. Aerosol absorption and size relationships showed a nonlinear dependence over the aerosol size ranges and allowed for determination of dominant absorbing aerosol types (urban/industrial and biomassburning). The main aerosol types are urban/industrial and biomass-burning, accounting for nearly 60% at two sites. But the ‘‘Mostly Dust’’ category accounted for about 2% at two sites. These relationships along with mean single scattering albedo spectra were used to categorize black carbon and dust as dominant absorbers and identify a third category where both black carbon and dust dominate absorption. As ˚ ngstro¨m exponents increased to 2.0, absorption A coarse mode particles became dominant. Further, single scattering albedo transitioned from spectra representing dust to urban/industrial and biomass˚ ngstro¨m exponents burning. As absorption A decrease to 1.25, the mean single scattering albedo values at Institute of Atmospheric Physics were less than those at Chinese Academy of Meteorological Sciences for all wavelengths. This could be related to the fact that Institute of Atmospheric Physics is a roadside site and with more black carbon aerosols. Strong absorptions were noted at 440 nm relative to longer wavelengths and higher values occurred at Institute of Atmospheric Physics for mean single scattering albedo and single scattering albedo with ˚ ngstro¨m exponent [0.8, and single extinction A scattering albedo with the fine mode fraction of extinction aerosol optical depth [0.66, indicating aerosols were dominated by fine mode particles at two sites and Institute of Atmospheric Physics had more absorbing particles. Strong absorption at 440 nm and strong scattering at longer wavelengths for single scattering albedo with absorption

123

˚ ngstro¨m exponents B0.8 and single scattering A albedo with the fine mode fraction of extinction aerosol optical depth \0.33 represented large dust particles. And the higher single scattering albedo values at Institute of Atmospheric Physics except at 440 nm may be associated with the dust absorbing particles and mixed particles. Acknowledgements This work was financially supported by the National Key R&D Program (2016YFA0601900), the NSFC Project of Nos. 41375153 and 41590874, CAMS Basic Research Project (2014R17), the Climate Change Special Fund of the CMA (CCSF201504), and Jiangsu Collaborative Innovation Center of Climate Change. Author Contributions Jie Yu and Huizheng Che designed and conceived the study. Hongbin Chen, Philippe Goloub, Quanliang Chen, Ke Gui, and Yu Zheng analyzed data. Hong Wang, Yaqiang Wang, Linchang An, Tianze Sun and Xiaoye Zhang significantly contributed to the manuscript’s revision. Jie Yu analyzed the data and wrote the paper. Compliance with Ethical Standards Conflicts of Interest The authors declare no conflict of interest.

References Aoki K, Fujiyoshi Y (2003) Sky radiometer measurements of aerosol optical properties over Sapporo, Japan. J Meteorol Soc Jpn 81:493–513. doi:10.2151/jmsj.81.49 Bi JR, Huang JP, Fu Q, Wang X, Shi JS, Zhang W, Huang ZW, Zhang BD (2011) Toward characterization of the aerosol optical properties over loess plateau of Northwestern China. J Quant Spectrosc Radiat 112:346–360. doi:10.1016/j.jqsrt.2010.09.006 Charlson RJ, Schwartz SE, Hales JM, Cess D, Coakley JA, Hansen JE (1992) Climate forcing by anthropogenic aerosols. Science 255:423–430. doi:10.1126/science.255.5043.423 Che HZ, Shi GY, Zhang XY (2007) Aerosol optical characteristics and its direct radiative forcing in Beijing. J Grad Sch Chin Acad Sci 24:699–704 Che H, Zhang X, Chen H, Damiri B, Goloub P, Li Z, Zhang X, Wei Y, Zhou H, Dong F, Li D, Zhou T (2009) Instrument calibration and aerosol optical depth validation of the China aerosol remote sensing network. J Geophys Res 114:D03206. doi:10.1029/2008JD011030 Che HZ, Wang YQ, Sun JY (2011) Aerosol optical properties at Mt. Waliguan observatory, China. Atmos Environ 45:6004–6009. doi:10.1016/j.atmosenv.2011.07.050 Che HZ, Xia XA, Zhu J, Li Z, Dubovik O, Holben B, Goloub P, Chen H, Estelles V, Cuevas-Agullo´ E et al (2014) Column aerosol optical properties and aerosol radiative forcing during a serious haze-fog month over North China plain in 2013 based on ground-based sunphotometer measurements. Atmos Chem Phys 14:2125–2138. doi:10.5194/acp-14-2125-2014 Che H, Zhang XY, Xia X, Goloub P, Holben B, Zhao H, Wang Y, Zhang XC, Wang H, Blarel L et al (2015a) Ground-based aerosol climatology of China: aerosol optical depths from the China Aerosol Remote Sensing Network (CARSNET) 2002–2013. Atmos Chem Phys 15:7619–7652. doi:10.5194/acp-15-76192015

Aerosol Sci Eng (2017) 1:78–92 Che HZ, Zhao HJ, Wu YF, Xia XA, Zhu J, Wang H, Wang YQ, Sun JY, Yu J, Zhang XY et al (2015b) Analyses of aerosol optical properties and direct radiative forcing over urban and industrial regions in Northeast China. Meteorol Atmos Phys 127:345–354. doi:10.1007/s00703-015-0367-3 Derimian Y, Karnieli A, Kaufman YJ, Andreae MO, Andreae TW, Dubovik O, Maenhaut W, Koren I (2008) The role of iron and black carbon in aerosol light absorption. Atmos Chem Phys 8:3623–3637. doi:10.5194/acp-8-3623-2008 Dong ZP, Li XM, Du CL, Zhang GJ (2013) Study on aerosol optical property in Xi’an Region. Plateau Meteorol 32:856–864. doi:10. 7522/j.issn.1000-0534.2012.00079 Dubovik O, King MD (2000) A flexible inversion algorithm for retrieval of aerosol optical properties from Sun and sky radiance measurements. J Geophys Res Atmos 105(D16):20673–20696. doi:10.1029/2000JD900282 Dubovik O, Smirnov A, Holben BN, King MD, Kaufman YJ, Eck TF, Slutsker I (2000) Accuracy assessments of aerosol optical properties retrieved from aerosol robotic network (AERONET) sun and sky radiance measurements. J Geophys Res 105:9791–9806. doi:10.1029/2000JD900040 Dubovik O, Holben B, Eck TF, Smirnov A, Kaufman YJ, King MD, Tanre D, Slutsker I (2002) Variability of absorption and optical properties of key aerosol types observed in worldwide locations. J Atmos Sci 59:590–608. doi:10.1175/15200469(2002)059\0590:VOAAOP[2.0.CO;2 Dubovik O, Sinyuk A, Lapyonok T, Holben BN, Mishchenko M, Yang P, Eck TF, Volten H, Mun˜oz O, Veihelmann B et al (2006) Application of spheroid models to account for aerosol particle nonsphericity in remote sensing of Desert dust. J Geophys Res. doi:10.1029/2005JD006619 Eck TF, Holben BN, Reid JS, Dubovik O, Smirnov A, O’Neill NT, Slutsker I, Kinne S (1999) Wavelength dependence of the optical depth of biomass buring, Urban, and Desert Dust Aerosols. J Geophys Res 104:31333–31349. doi:10.1029/1999JD900923 Eck TF, Holben BN, Dubovik O, Smirnov A, Goloub P, Chen HB, Chatenet B, Gomes L, Zhang XY, Tsay SC et al (2005) Columnar aerosol optical properties at AERONET sites in Central Eastern Asia and Aerosol transport to the tropical MidPacific. J Geophys Res. doi:10.1029/2004jd005274 Eck TF, Holben BN, Reid JS, Sinyuk A, Dubovik O, Smirnov A, Giles D, O’Neill NT, Tsay SC, Ji Q et al (2008) Spatial and temporal variability of column-integrated aerosol optical properties in the Southern Arabian Gulf and United Arab Emirates in summer. J Geophys Res. doi:10.1029/2007JD008944 Eck TF, Holben BN, Sinyuk A, Pinker RT, Goloub P, Chen H, Chatenet B, Li Z, Singh RP, Tripathi SN et al (2010) Climatological aspects of the optical properties of fine/coarse mode aerosol mixtures. J Geophys Res 115:D19205. doi:10. 1029/2010JD014002 Eck TF, Holben BN, Reid JS, Giles DM, Rivas MA, Singh RP, Tripathi SN, Bruegge CJ, Platnick S, Arnold GT et al (2012) Fog- and cloud-induced aerosol modification observed by the Aerosol Robotic Network (AERONET). J Geophys Res. doi:10. 1029/2011JD016839 Giles DM, Holben BN, Tripathi SN, Eck TF, Newcomb WW, Slutsker I, Dickerson RR, Thompson AM, Mattoo S, Wang SH et al (2011) Aerosol properties over the Indo-Gangetic plain: a mesoscale perspective from the TIGERZ experiment. J Geophys Res. doi:10.1029/2011JD015809 Giles DM, Holben BN, Eck TF, Sinyuk A, Smirnov A, Slutsker I, Dickerson RR, Thompson AM, Schafer JS (2012) An analysis of AERONET aerosol absorption properties and classifications representative of aerosol source regions. J Geophys Res. doi:10. 1029/2012JD018127

91 Gyawali M, Arnott WP, Lewis K, Moosmu¨ller H (2009) In situ aerosol optics in Reno, NV, USA during and after the summer 2008 California wildfires and the influence of absorbing and nonabsorbing organic coatings on spectral light absorption. Atmos Chem Phys 9:8007–8015. doi:10.5194/acp-9-8007-2009 Hansen J, Sato M, Ruedy R, Lacis A, Oinas V (2000) Global warming in the twenty-first century: an alternative scenario. Proc Natl Acad Sci 97:9875–9880. doi:10.1073/pnas.170278997 Holben BN, Eck TF, Skitsker I, Tanre´ D, Bziis JP, Setzer A, Vermote E, Reagan JA, Kazlfnzan YJ, Nnknjima T et al (1998) AERONET-A federated instrument network and data archive for aerosol characterization. Remote Sens Environ 66:1–16. doi:10.1016/S0034-4257(98)00031-5 IPCC (2013) Climate change 2013: the physical science basis. Contribution of working group I to the fifth assessment report of the intergovernmental panel on climate change. Cambridge University Press, Cambridge Li W, Li P, Sun G, Zhou S, Yuan Q, Wang W (2011) Cloud residues and interstitial aerosols from non-precipitating clouds over an industrial and urban area in northern China. Atmos Environ 45:2488–2495. doi:10.1016/j.atmosenv.2011.02.044 Myhre G (2009) Consistency between satellite-derived and modeled estimates of the direct aerosol effect. Science 325:187–190. doi:10.1126/science.1174461 Prasad AK, Singh RP (2007) Changes in aerosol parameters during major dust storm events (2001–2005) over the Indo-Gangetic Plains using AERONET and MODIS data. J Geophys Res. doi:10.1029/2006JD007778 Qi B, Hu DY, Che HZ, Du RG, Wu YF, Xia XA, Zha B, Liu J, Niu YW, Wang H, Zhang XY, Shi GY (2016) Seasonal variation of aerosol optical properties in an urban site of the Yangtze Delta region of China. Aerosol Air Qual Res 16:2884–2896. doi:10. 4209/aaqr.2015.05.0350 Ramanathan V, Crutzen PJ, Kiehl JT, Rosenfeld D (2001) Aerosols, climate, and the hydrological cycle. Science 294:2119–2124. doi:10.1126/science.1064034 Smirnov A, Holben BN, Eck TF, Dubovik O, Slutsker I (2000) Cloud screening and quality control algorithms for the AERONET database. Remote Sens Environ 73:337–349. doi:10.1016/ S0034-4257(00)00109-7 Song SJ, Wua Y, Xu JY, Ohara T, Hasegawa SC, Li JQ, Yang L, Hao JM (2013) Black carbon at a roadside site in Beijing: temporal variations and relationships with carbon monoxide and particle number size distribution. Atmos Environ 77:213–221. doi:10. 1016/j.atmosenv.2013.04.055 Sun YL, Du W, Fu PQ, Wang QQ, Li J, Ge XL, Zhang Q, Zhu CM, Ren LJ, Xu WQ et al (2016) Primary and secondary aerosols in Beijing in winter: sources, variations and processes. Atmos Chem Phys 16:8309–8329. doi:10.5194/acp-16-8309-2016 Toledano C, Wiegner M, Groß S, Freudenthaler V, Gasteiger J, Mu¨ller D, Mu¨ller T, Schladitz A, Weinzierl B, Torres B, O’neill NT (2011) Optical properties of aerosol mixtures derived from sun-sky radiometry during SAMUM-2. Tellus B 63:635–648. doi:10.1111/j.1600-0889.2011.00573.x Twomey SA, Piepgrass M, Wolfe TL (1984) An assessment of the impact of pollution on the global cloud albedo. Tellus 36B:356–366. doi:10.1111/j.1600-0889.1984.tb00254.x Wang P, Che HZ, Zhang XC, Song QL, Wang YQ, Zhang ZH, Dai X, Yu DJ (2010a) Aerosol Optical Properties of Regional Background Atmosphere in Northeast China. Atmos Environ 44:4404–4412. doi:10.1016/j.atmosenv.2010.07.043 Wang X, Huang JP, Zhang RD, Chen B, Bi JR (2010b) Surface measurements of aerosol properties over Northwest China during ARM China 2008 deployment. J Geophys Res 115:D00K27. doi:10.1029/2009JD013467

123

92 Xiao ZY, Jiang H, Chen J, Wang B, Jiang ZS (2011) Monitoring the aerosol optical properties over Hangzhou using remote sensing data. Acta Sci Circumst 31:1758–1767 Xie Y, Zhang Y, Xiong XX, Qu JJ, Che HZ (2011) Validation of modis aerosol optical depth product over China using Carsnet measurements. Atmos Environ 45:5970–5978. doi:10.1016/j. atmosenv.2011.08.002 Yan H, Wu Y, Zhang SJ, Song SJ, Fu LX, Hao JM (2014) Emission characteristics and concentrations of vehicular black carbon in a typical freeway traffic environment of Beijing. Acta Sci Circum 34:1891–1899. doi:10.13671/j.hjkxxb.2014.0523 Zege EP, Ivanov AP, Katsev IL (1991) Image transfer through a scattering medium. Springer, New York Zhang XY, Wang YQ, Niu T, Zhang XC, Gong SL, Zhang YM, Sun JY (2012) Atmospheric aerosol compositions in China: spatial/ temporal variability, chemical signature, regional haze

123

Aerosol Sci Eng (2017) 1:78–92 distribution and comparisons with global aerosols. Atmos Chem Phys 12:779–799. doi:10.5194/acp-12-779-2012 Zhang JK, Sun Y, Liu ZR, Ji DS, Hu B, Liu Q, Wang YS (2013a) Characterization of submicron aerosols during a serious pollution month in Beijing (2013) using an aerodyne high-resolution aerosol mass spectrometer. Atmos Chem Phys 13:19009–19049. doi:10.5194/acpd-13-19009-2013 Zhang R, Jing J, Tao J, Hsu SC, Wang G, Cao J, Lee CSL, Zhu L, Chen Z, Zhao Y, Shen Z (2013b) Chemical characterization and source apportionment of PM2.5 in Beijing: seasonal perspective. Atmos Chem Phys 13:7053–7074. doi:10.5194/acp-13-7053-2013 Zhao HJ, Che HZ, Zhang XY, Ma YJ, Wang YF, Wang XX, Liu C, Hou B, Che HC (2013) Aerosol optical properties over urban and industrial region of Northeast China by using ground-based sunphotometer measurement. Atmos Environ 75:270–278. doi:10. 1016/j.atmosenv.2013.04.048